[Subsurface-trac] #486: Manually entered profile not transferred correctly to divelogs.de]
willemferguson at zoology.up.ac.za
Mon Apr 7 02:51:06 PDT 2014
In the recreational field the assumption of a constant interval among
samples is, as far as I know, valid.
However, there are computers that record information only when there has
been a change to be recorded. So, for instance, if the temperature has
not changed, no temperature is recorded. The Poseidon, as an example,
only triggers a logging event when a change in conditions has occurred.
And the Poseidon is mostly a recreational machine.
The bottom line is that, in the longer term, one needs to start making
provision for equipment that does not log at regular intervals. Yes, its
painful to provide for a flexible system, but that is the price to pay
to support a wide array of hardware and users. And this is what sets
meritorious open source systems (such as divelogs.de) separate from
vendor-provided software which does not have the adaptability to store
information from the diverse hard gear systems that a single diver uses.
As technical diving (equipment and training) becomes more and more
accessible, this need for diversity will only increase in future.
On 07/04/2014 11:20, Rainer Mohr wrote:
> Am 07.04.14 07:32, schrieb Miika Turkia:
>> Rainer - is that fixed sample interval format the only format
>> divelogs.de <http://divelogs.de> understands? Or is there another
>> format where we could give
>> the time stamp in the sample as we do in Subsurface?
> I'm afraid so, yes. I save the sample interval with the dive's data
> and the samples have only the depth and not when it occurs.
> Since Dive computers have regular samples every x seconds, there was
> no need until now to make this variable. Changing this would hurt on
> my side :-(
>> If not, I guess we'd have to create synthetic samples to try to get
>> "reasonably close" (in this case we could get it nearly perfectly).
>> Miika, would that be doable in your xslt or would we need to
>> things to have the correct samples?
>> It would be great to have the timestamp option, but I should be able
>> to generate synthetic samples in XSLT as well. What is the time wise
>> granularity of the manually added samples? 1 minute as the sample
>> dive in bug report? Something larger than 1 second would be good
>> here, as then it is easy to use that as sample interval and not end
>> up with insanely large export file.
> That would be great and I wouldn't have to change my database structure...
> You can use anything you like as sample interval, I guess the smallest
> difference (in seconds) between two samples in the subsurface dive
> makes sense. Then use synthetic samples for the larger intervals. If
> that's feasible, I'd be greatly happy.
> If it's problematic on your side, I can try to insert fake samples on
> my side too, if there is no sample interval defined in the XML I get,
> but only timestamps...
> Your shot, let me know if you want me to do it.
> subsurface mailing list
> subsurface at hohndel.org
Director: Centre for Environmental Studies,
University of Pretoria
Tel: 012-4202017 / 012-4204048
e-mail: willemferguson at zoology.up.ac.za
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the subsurface